academic-figure-draft-analyzer

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted content including academic manuscripts, PDF files, and web pages. This creates a potential surface for indirect prompt injection where malicious instructions hidden in these documents could attempt to influence the agent's behavior.
  • Ingestion points: The skill accepts manuscript text, outlines, local PDF files, and content extracted from academic paper URLs (SKILL.md).
  • Boundary markers: The instructions do not define explicit delimiters or instructions for the agent to ignore or isolate potentially malicious embedded commands within the ingested source material.
  • Capability inventory: The skill's primary capability is the generation of structural figure plans and strategy recommendations; it does not currently specify scripts for arbitrary shell command execution or direct network exfiltration.
  • Sanitization: The workflow does not include specific validation or sanitization steps to filter out natural language instructions embedded in the documents being processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 03:33 AM
Security Audit — agent-trust-hub — academic-figure-draft-analyzer